764 TITRATION OF TOXINS AND ANTITOXINS 



have been observed by Ramon (1923), Povitzky (1927), and others. A serum with a 

 high titre of agglutinin may have neither antitoxic nor flocculating power, and a serum 

 devoid of agglutinins may be both antitoxic and capable of causing flocculation. Ag- 

 glutinins are demonstrable in the supernatant fluid after flocculation, and the anti- 

 toxic content may or may not be reduced by absorption of agglutinins from the serum. 

 Moloney and Weld (1926) have shown that diphtheria toxin-antitoxin flocculation is 

 not group-specific according to the serological relationship of the bacterial strains. In 

 this respect, the flocculation results accord with those of animal tests in their indica- 

 tion of the serological unity of diphtheria toxins. 



The precipitate which forms when antibacterial precipitin is added to diphtheria 

 toxin does not remove the toxin nor interfere with its subsequent flocculation with an- 

 titoxin. A new flocculation occurs when this toxin, from which precipitable bacterial 

 residues have been removed, is added in equivalent proportions to antitoxin. After 

 the removal of the precipitate formed by antitoxin in an "indicating tube," no further 

 flocculation occurs when more antitoxin is added. In repeating these experiments, I 

 have confirmed the results obtained by H. Schmidt (1926), demonstrating the inde- 

 pendence of the toxin-antitoxin flocculation with reference to antibacterial precipitins. 



Other controls, such as mixtures of toxins with normal sera, and mixtures of toxins 

 with unrelated antitoxins, have been uniformly negative, furnishing additional evi- 

 dence of the specificity of the toxin-antitoxin flocculation reaction. 



The precipitate formed during flocculation is composed of toxin and antitoxin, 

 in a dissociable union from which both substances can be recovered by heat in a 

 slightly acid or alkaline solution, or by the use of solutions of Nal — Calmette and 

 Massol (1909), Ramon (1923), and Madsen and S. Schmidt (1926). The flocculation 

 reaction, therefore, provides a means by which these substances can be isolated in a 

 state of purity hitherto not obtained — Locke and Main (1926). Ramon (1923) ob- 

 tained a solution containing 60,000 units of antitoxin per gram of dried material from 

 treatment of this precipitate. Abt and Erber (1926) found that the weight of the pre- 

 cipitate increased in proportion to the number of antitoxic units neutralized in mix- 

 tures of tetanus toxin and antitoxin. Chemical analyses of the precipitate have been 

 made by Flossner and Kutscher (1924) and by Hartley (1925, 1926). The toxin-anti- 

 toxin precipitate contained protein, lipoids, and ammonium magnesium phosphate. 

 Hartley found from 0.00134 to 0.00203 mgm. nitrogen per unit of antitoxin in the 

 precipitate from balanced mixtures of diphtheria toxin and antitoxin. These precipi- 

 tates, as shown by Hartley (1925, 1926) and by H. Schmidt (1925), are antigenic and 

 can be used for the immunization of man and lower animals. Hartley found that an 

 amount of precipitate containing o.ooi mgm. N was sufficient to immunize a guinea 

 pig. Precipitates from neutral mixtures were non-toxic and moderately good antigens. 

 Very little antigenic toxin remained in the supernatant fluid in a neutral flocculating 

 mixture. Precipitates from underneutralized mixtures were slightly toxic and very 

 active antigenically. Precipitates from overneutralized mixtures were non-toxic and 

 had small antigenic value. The antigenic properties of these precipitates, their corre- 

 spondence in weight to units neutralized and their composition reflecting the constitu- 

 tion of the mixtures from which they are derived, strengthen the evidence in favor 

 of the specificity of the toxin-antitoxin flocculation reaction. 



